Item

Abstract

Chemical buoyancy within the lithosphere may be prevalent in certain depth regions. To constrain those depths, we follow an approach similar to Wang et al. (2023) and plot the difference between residual topography—where only crustal isostatic topography has been removed—and dynamic topography—where only sub-lithospheric density anomalies are considered—against lithosphere thickness. In theory, this difference should be caused by density anomalies in the lithosphere, and the slope of the fitting line versus lithosphere thickness should hence indicate depth-dependent density anomalies in the lithosphere. When computing dynamic topography, lithosphere thickness and density anomalies outside the continental lithosphere are derived from tomography. Within the continental lithosphere, density anomalies are set to a reference value, for which we use either the global mean, or the depth-dependent mean values below mid-ocean ridges. With densities in the continental lithosphere set to global mean, we typically find a break in slope around 150 km thickness, from nearly zero above to a negative slope below. With densities set to mid-ocean ridge values, there is still a break in slope but the slope above 150 km is also negative. This indicates that chemical density anomalies that cause lithospheric buoyancy are concentrated in the upper ≈150 km. Whether or not there are substantial density anomalies also above 150 km mainly depends on whether the global mean, or mid-ocean ridges are used for reference.